Together AI introduces Provisioned Throughput for guaranteed open model inference

Together AI has launched Provisioned Throughput, a new inference form factor that provides reserved capacity for frontier open models with token-based pricing and a 99% uptime SLA. This option bridges the gap between best-effort serverless inference and complex dedicated GPU management by offering predictable costs and guaranteed availability.

• Available today for MiniMax M3 and GLM-5.2 in North America, EMEA, and beyond with a one-month minimum term.
• Customers buy Provisioned Throughput Units (PTUs) priced at $0.05 per PTU per minute to secure exclusive capacity.
• Costs run up to 90% below Claude Opus 4.8 list price, with MiniMax M3 PTUs delivering roughly $0.36 per million input tokens and $2.16 per million output tokens at full utilization.
• The API is identical across Serverless, Dedicated Inference, and Provisioned Throughput, allowing seamless integration.

This release enables companies to migrate production workloads from proprietary APIs to open models with the reliability guarantees required for business-critical applications.